Merge remote-tracking branch 'origin/master'

pull/5/head
Thalles 2020-03-14 18:33:54 -03:00
commit b1e8f2c3d9
2 changed files with 6 additions and 5 deletions

View File

@ -72,7 +72,7 @@ Feature evaluation is done using a linear model protocol.
Features are learned using the ```STL10 train+unsupervised``` set and evaluated in the ```test``` set;
Check the [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/sthalles/SimCLR/blob/9d071bb3dd93e921217c415cf0924aad2d0b13eb/feature_eval/linear_feature_eval.ipynb) notebook for reproducibility.
Check the [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://github.com/sthalles/SimCLR/blob/master/feature_eval/linear_feature_eval.ipynb) notebook for reproducibility.
| Linear Classifier | Feature Extractor | Architecture | Feature dimensionality | Projection Head dimensionality | Epochs | STL10 Top 1 |

View File

@ -58,10 +58,11 @@
"colab": {}
},
"source": [
"folder_name = 'Mar13_22-46-30_thallessilva'\n",
"folder_name = 'Mar14_05-52-52_thallessilva'\n",
"\n",
"!gdown https://drive.google.com/uc?id=12kKgvo4h41G9qnDdhDnZXFlR5_aqvaVR\n",
"!unzip Mar13_22-46-30_thallessilva.zip\n",
"# !gdown https://drive.google.com/uc?id=12kKgvo4h41G9qnDdhDnZXFlR5_aqvaVR # ResNet 18 --> 40 epochs trained\n",
"!gdown https://drive.google.com/uc?id=1LjuZ1RmhotrnugprRQc2Exk0EbQHMJhL # ResNet 18 --> 80 epochs trained\n",
"!unzip Mar14_05-52-52_thallessilva\n",
"!ls"
],
"execution_count": 0,
@ -406,7 +407,7 @@
{
"cell_type": "code",
"metadata": {
"id": "Ugf0L7rKLjRd",
"id": "fXy_YX8_b7gL",
"colab_type": "code",
"colab": {}
},